Many times during development, the project was on the brink of being cancelled by the FIEA faculty. One of the main things that finally saved BizarreCraft from cancellation was its prototype. For this reason, the prototype deserves an exclamation point.
For weeks, the designers struggled to try and figure out how to make our RTS game different from any other RTS. Many ideas were tossed around, but there was no way to tell what could work and be fun. We could debate the issue for hours on end, but, in the end, the only real way to tell if a system works is to experience it.
This was not going to be feasible. Our programmers were already overwhelmed with the amount of work required to just get the basic features of an RTS functional. There was no way they could complete all of these features by the time a final decision needed to be made. As such, they would never get to these experimental features the designers were dreaming up so they could test them out early in development.
The lack of an answer as to how our RTS would be different from all other RTS was holding up too much of development. We were close to having to either make an RTS clone or commit to a set of features that could ensure the downfall of our project.
It was our programmers' love of RTS games that led to the great breakthrough that was the BizarreCraft prototype. During the coursework at FIEA that we were completing in parallel with development, each student was assigned to use a consumer grade level editor and build a game level.
Some of our programmers chose to use the Warcraft III level editor. As they worked on their levels and became more familiar with the editor, they began to see the powerful potential the editor posed. Upon completing their assignment, they proposed to the team to test out the design game ideas by simulating them through the Warcraft III editor. At first, the designers were not sure this would give them the flexibility they hoped for, but, after days exploring the editor, they were pumping out new simulations every couple hours.
Eventually they put together a prototype of what would eventually become BizarreCraft. This allowed them to not only prove the entertainment value to themselves, but also to the rest of the team. Months before the team could play a functional version of BizarreCraft, they were having a blast playing its prototype. This also allowed the team to show the faculty members who had pushed for our cancellation that we were headed in the direction of a great game.
As great as the prototype was for the team and for development, it came with an unexpected cost. Even with this cost, I do not think anyone on the team would do things differently in terms of the prototype. I will discuss this cost later in this postmortem.
4. Characters, Not Units
You cannot have an RTS game without units to command. It is a staple of the genre, even if the number of unit you control is one. We wanted our artists to be able to have a great portfolio piece with some memorable units. We also felt having many unique units would enhance the strategic portion of our game.
Early on, the units we were going to use in the game were coming in and leaving from development like they were on a carousel. We could not commit to any units. We could not figure out what the units would do. We did not know what the units would look like.
We could not come up with a unified look for the units. We felt helpless and without direction. We did not want our units to be the nondescript peons found in many other RTS games. We wanted them to be memorable characters. And that was it.
We had to stop looking at creating units with a set of stats, but on creating characters. We needed to create characters to inhabit our world. They needed personalities. We needed to think what they would do if they were a character in the humorous story of an absurd army.
In doing this, we developed an in-depth story for our world and for each character. The characters suddenly telling us what their attacks would be based on their personalities and back story. Once we had the fictional side of the character complete, we could fill in the unit side of things. This mentality allowed us to create memorable characters.
We created a hybrid monkey and ostrich that ran around like crazy screaming at anyone nearby. We created a two-legged wrecking ball that didn't like to hurt people, so he used his wrecking ball as a pendulum to hypnotize foes. It felt like we had opened the floodgates of creativity. We converted these fictional stories into statistical data.
We cut characters whose stats unbalanced the game, we tweaked stories and stats for characters to make sure each side had comparable tools at their disposal. Only downside to this methodology is that people get attached to characters. No one cares about units. We did not have units anymore, which meant we had created characters that the team would be attached to. Things that people are attached to are always harder to cut.